Rainfall data restoration method and system, flood forecasting method and device and medium

Abstract

The invention discloses a rainfall data restoration method and system, a flood forecasting method and device and a medium, and the method comprises the steps: S1, reading the original monitoring data of all rainfall stations in a target region, screening out low-health-degree stations with abnormal data or monitoring failure, and determining the precise space coordinates of the low-health-degree stations; s2, screening out stations of which the health degree reaches a predetermined standard, and calculating the spatial distance between the target station and each adjacent station; s3, calculating the contribution weight of each adjacent site to the data restoration of the target site; s4, interpolating missing or abnormal data of the low-health-degree stations according to the rainfall data of the adjacent stations and the weights of the adjacent stations to generate continuous and high-precision rainfall time sequence data; and S5, performing statistical test and continuity evaluation on the data, and eliminating abnormal segments with non-ideal interpolation effects. According to the method, spatial correlation is fully mined, and the integrity and reliability of rainfall data can be improved, so that accurate matching of data among stations is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of hydrological data processing and flood forecasting, and particularly relates to a method and system for repairing abnormal rainfall data based on inverse distance weighted interpolation, a flood forecasting method, device and storage medium integrating a flood forecasting model. Background Art

[0002] The accuracy and continuity of current rainfall monitoring data are of crucial significance for flood forecasting and hydrological analysis. However, due to factors such as sensor failures, complex installation environments, and data transmission interruptions, rainfall stations often have missing or abnormal data, which directly affects the accuracy and reliability of subsequent flood forecasting models.

[0003] Traditional data repair methods mainly rely on linear interpolation, nearest neighbor interpolation, or simple mean completion techniques. However, these methods do not fully consider the non - homogeneity and local characteristics of rainfall data in spatial distribution, and often struggle to handle local rainfall variations under complex terrains and extreme climate conditions.

[0004] In recent years, with the development of geographic information systems (GIS), remote sensing technology, and spatial statistical methods, the inverse distance weighted (IDW) interpolation method has received extensive attention due to its simplicity, intuitiveness, and ease of implementation. The IDW method uses data from neighboring stations and assigns different weights according to the spatial distance between each station and the target station, thereby achieving adaptive repair of missing or abnormal data.

[0005] However, the repair effect of the IDW method depends to a large extent on the reasonable selection of parameter settings (such as distance power exponent and small constant), and the balance of the spatial distribution of monitoring stations. Therefore, it cannot accurately repair low - health stations using information from high - health stations under conditions of sparse or uneven data distribution. In addition, with the increasing requirement for the accuracy of input data in flood forecasting, existing data repair technologies urgently need to be organically integrated with flood forecasting models to build a complete closed - loop system from data repair to early warning decision - making.

[0006] Therefore, this application specifically proposes a method for repairing rainfall data to solve the above - mentioned technical problems. Summary of the Invention

[0007] The main purpose of the present invention is to provide a method for repairing rainfall data, which can not only fully explore spatial correlation, improve the integrity of rainfall data, but also be seamlessly integrated into a flood forecasting system, and has important theoretical and practical significance for improving the ability of disaster prevention and reduction and the level of water resource management, so as to solve the technical problems proposed in the background art.

[0008] The present invention adopts the following technical solutions to solve the above - mentioned technical problems: A method for repairing rainfall data, comprising the following steps: S1. Read the original monitoring data of all rainfall stations in the target area and record it as a time series where is the number of all rainfall stations in the target area, represents the rainfall value at time . At the same time, obtain the health index and geographical location information coordinates of each station, where is the latitude of the station, is the longitude of the station, and based on a preset threshold screen out the low-health stations with abnormal data or monitoring failures and determine their precise spatial coordinates; S2. Screen out the stations with health reaching the predetermined standard within the preset geographical radius, and calculate the spatial distance between the target station and each neighboring station using the spherical distance formula; S3. According to the inverse distance weighting principle, calculate the contribution weight of each neighboring station to the data repair of the target station using the spherical distance, power exponent and a small constant; S4. Adopt the IDW interpolation method to interpolate the missing or abnormal data of the low-health stations with the rainfall data and weights of the neighboring stations to generate continuous and high-precision rainfall time series data; S5. Conduct statistical tests and continuity evaluations on the data, and eliminate the abnormal segments with unsatisfactory interpolation effects.

[0009] Preferably, the specific operation process of step S2 includes: S21. Neighboring station screening. For each low-health station , set a fixed geographical radius around the target station to determine the candidate high-health stations. Define the candidate station set as:

[0010] where represents the spatial distance between station and station .

[0011] S22. Calculate the distance between stations. The calculation method is as follows:

[0012] where is the radius of the earth, is the difference in latitude between the two stations, is the difference in longitude between the two stations, and are the latitude and longitude of the site respectively, and are the latitude and longitude of the site respectively.

[0013] Convert the coordinates with the coordinates of the candidate high-health sites There is an interpolation calculation formula as follows:

[0014]

[0015]

[0016] Substitute the above interpolation into the distance calculation formula between sites, and we get:

[0017] to obtain the corresponding set for each site in the set and its distance ; wherein, and are the radian coordinates after conversion of the latitude coordinates and respectively, and are the radian coordinates after conversion of the longitude coordinates and respectively. Preferably, the S3 step and the S4 step are used to calculate the weights and generate continuous and high-precision rainfall time series data, and we have: For each high-score neighboring site , calculate the weight , and the calculation formula is:

[0018] wherein, is used to ensure that there is no division-by-zero error when the site approaches zero, making the closer sites have larger weights, is the distance between each site and its adjacent site is the total number of high-score neighboring sites, is the starting number of the summation, is the weight adjustment value; By using the IDW interpolation method, based on the calculated weights ​Calculate the target site At time The rainfall estimate value , there is:

[0019] Used for high-precision interpolation using adjacent site data in the case of low rainfall site health or data loss, generating complete rainfall time series data, Is the total number of high-score adjacent sites, Is the site At time The rainfall estimate value

[0020] Preferably, the specific operation process of the S5 step includes: S51. Set the repair data Of the target site Within the time interval As , at this time for the repair data, calculate the mean value And variance , the calculation formula is:

[0021] Among them, Is the time interval length; S52. Set the continuity error index , used to evaluate the data continuity error index The calculation formula is:

[0022] When the continuity error Exceeds the preset threshold , eliminate the interpolation within this time interval

[0023] A rainfall data repair system for performing any of the above-mentioned rainfall data repair methods, including: The site data acquisition module is used to collect the original monitoring data of all rainfall sites in the target area and obtain the health index and geographical location information; The data calculation and repair module is used to perform interpolation calculation on the missing or abnormal data of low-health sites based on the rainfall data of adjacent sites and their weights to generate continuous and high-precision rainfall time series data; The model verification module is used to perform statistical tests and continuity evaluations on the repair data, eliminate abnormal segments with unsatisfactory interpolation effects, and fuse the verified data with other hydrological monitoring data as input features of the flood forecasting model.

[0024] ​A flood forecasting method, based on the rainfall data repair method described in any of the above, obtains the repaired rainfall data, and has the following specific operation steps: L1. Based on the rainfall estimates at each station , construct a flood forecasting model, fuse the verified data with other hydrological monitoring data, and use them as the input features of the flood forecasting model; L2. Quantitatively evaluate the forecasting model using preset indicators including the mean squared error, and optimize the data repair algorithm and model parameters based on the feedback to construct an adaptive early warning system; L3. Perform flood prediction operations through the flood forecasting model.

[0025] Preferably, the fusion method of the input features in step L1 includes: Select the repaired data that passes the statistical test and continuity evaluation , and fuse it with other hydrological monitoring data to construct a joint input feature vector , there is:

[0026] Among them, , represent the reservoir water level data and flow data respectively, represents the effective time period.

[0027] Preferably, the specific optimization process in step L2 includes: L21. Quantitatively evaluate the output of the flood forecasting model using specified indicators including the mean squared error. Within the time interval length , the quantitative evaluation with the mean squared error is defined as:

[0028] Among them, is the hydrological index including the true observed rainfall data, is the model prediction value; L22. According to the calculated value, after comparing with the preset performance indicators, tune the hyperparameters of the data repair algorithm and the forecasting model, and use gradient descent to minimize the model output error, there is: , where represents the set of all parameters to be optimized; L23. Construct an adaptive early warning system for real-time monitoring and feedback adjustment of the model output error. When exceeds the specified early warning threshold , the system automatically triggers parameter re-optimization or an alarm mechanism.

[0029] In another aspect, the present invention also discloses a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to execute the steps of the above method.

[0030] In yet another aspect, the present invention also discloses a computer device comprising a memory and a processor, where the memory stores a computer program, which, when executed by the processor, causes the processor to execute the steps of the above method.

[0031] As can be seen from the above technical solutions, the present invention provides a method for repairing rainfall data. Compared with the prior art, the present invention has the following advantages: 1. By performing time alignment, missing value filling, and weighted fusion of neighboring stations on the rainfall station data, the present invention significantly improves the integrity and reliability of rainfall data, ensures accurate matching of data among monitoring stations, and thus provides high-quality input data for flood forecasting models.

[0032] 2. By introducing the inverse distance weighted interpolation (IDW) method during the rainfall data repair process, the present invention can make full use of the spatial correlation of neighboring stations with high health levels, effectively making up for the limitations of traditional simple interpolation methods in data-sparse regions, and thus significantly improving the repair accuracy and continuity of data for low-health-level stations, facilitating the provision of high-quality, high spatio-temporal matching input data for flood forecasting models.

[0033] 3. By using the spherical distance calculation formula to accurately describe the geographical spatial relationship between stations during the neighboring station screening stage, and leveraging the spatial correlation and data redundancy among multiple stations, the present invention can overcome the errors in traditional planar distance calculations, enhance the topographic adaptability of interpolation results, ensure the accuracy and continuity of rainfall data repair, and further enhance the reliability of rainfall data repair under complex terrain conditions, being applicable to fields such as flood forecasting, flood control scheduling, and hydrological monitoring.

[0034] 4. By introducing a statistical test and continuity error assessment mechanism during the data verification process, the present invention can automatically eliminate abnormal data segments with unsatisfactory interpolation effects, thus ensuring the consistency and continuity of the repaired data in a statistical sense, and ultimately avoiding the risk of misjudgment of flood forecasting models caused by local interpolation biases.

[0035] 5. By achieving spatio-temporal fusion of multi-source hydrological data during the model integration stage, the present invention can construct a joint input feature vector to meet the requirements of flood forecasting models, so as to enhance the dynamic response ability of the model to multi-dimensional hydrological processes, and ultimately enhance the spatio-temporal resolution and prediction accuracy of flood forecasting results.

[0036] 6. By introducing an MSE-driven adaptive adjustment mechanism in the model optimization stage, the present invention can provide real-time feedback and optimize the data repair algorithm and model parameters, thereby dynamically enhancing the system's robustness against extreme climate events and emergencies, facilitating the ultimate construction of an intelligent flood prevention early warning system with self-adaptive capabilities.

[0037] 7. Through the organic combination of spatial correlation modeling, multi-source data fusion, and self-adaptive optimization techniques, the present invention can break through the limitations of traditional data repair methods, thereby comprehensively improving the data quality, prediction accuracy, and real-time response capabilities of the flood forecasting system. Ultimately, it can provide scientific and reliable technical support for flood control scheduling, water conservancy project management, and disaster prevention and control, enhancing the overall accuracy and real-time response capabilities of the forecasting system.

[0038] It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of the present invention, nor is it used to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. Of course, it is not necessary for any product implementing the present invention to achieve all the above-mentioned advantages simultaneously. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The accompanying drawings forming a part of this application are used to provide a further understanding of the present invention. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation to the present invention. In the drawings: Figure 1 is a schematic diagram of the overall process of the present invention; Figure 2 is a schematic diagram of the spatial distribution and weight calculation of the target low-health sites and their neighboring high-health sites of the present invention; Figure 3 is a schematic diagram of the time-series comparison of rainfall data before and after repair of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0040] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Without conflict, the embodiments in this application and the features in the embodiments can be combined with each other. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0041] In the embodiments, refer in detail to Figures 1 to 3 .

[0042] Such as Figure 1As shown in the figure. The rainfall data repair method and the flood forecasting method of the integrated flood forecasting model proposed in the embodiments of the present invention are used to integrate the repaired data into the flood forecasting model to improve the forecasting accuracy.

[0043] The rainfall data repair method first collects and preprocesses the original monitoring data of all rainfall stations in the target area, and at the same time obtains the health index and geographical location information corresponding to each station, and identifies the low-health stations according to the health identification rules; subsequently, it screens out neighboring stations with a health index greater than the threshold within the set geographical radius, and calculates the distance between the target station and each neighboring station using the spherical distance formula; then, according to the inverse distance weighted principle, calculates the contribution weight of each neighboring station to the data repair of the target station, where a power exponent and a small constant are introduced to ensure numerical stability; uses the rainfall data of the neighboring stations and the calculated weights to repair the data of the low-health stations by the IDW interpolation method to generate continuous and high-quality rainfall time series data, which is used as the rainfall distribution basis for the flood forecasting model; subsequently, conducts statistical tests and continuity evaluations on the generated rainfall time series data, and eliminates abnormal segments with unsatisfactory interpolation effects; finally, uses the verified rainfall time series data as input features, integrates it into the flood forecasting model during the operation of the flood forecasting method, and evaluates the model performance and optimizes the parameters using indicators such as mean square error to form an adaptive early warning system.

[0044] It should be noted that in the specific embodiments of the present invention, any forecasting model such as the SWAT (Soil&Water Assessment Tool) model, the HEC-HMS (Hydrologic Modeling System) model, or the Xin'anjiang model can be selected as the flood forecasting model to implement the above data fusion and forecasting process; in addition, in the subsequent embodiments of this application, the Xin'anjiang model is taken as an example.

[0045] Furthermore, the overall method of rainfall station data repair and flood forecasting model integration based on inverse distance weighted interpolation first collects original monitoring data from multiple rainfall monitoring stations, including the rainfall monitoring records, health indexes, and geographical location information of each station; the rainfall monitoring data reflects the rainfall values of each station at different time points, the health index is used to evaluate the quality and reliability of the monitoring data, and the geographical location information includes the longitude and latitude of the station. At this time, by sorting the above various types of data in time, aligning the data, and filling in missing values, the consistency, integrity, and accuracy of the input data are effectively improved, providing scientific and reliable data support for flood forecasting, and having high engineering adaptability and actual promotion prospects.

[0046] The specific implementation process includes the following steps: Step S1: Data collection, preprocessing, and identification of low-health stations. Read the original monitoring data of all rainfall stations in the target area. Meanwhile, obtain the health index and geographical location information of each station, and screen out the low-health stations with abnormal data or monitoring failures according to the preset threshold to determine their precise spatial coordinates.

[0047] S11. Data collection and preprocessing. For each rainfall station ( , being the number of all rainfall stations in the target area), collect the original monitoring data, denoted as the time series , where represents the rainfall value at time .

[0048] S12. Extract the health index of each station from the original data (calculated through indicators such as equipment status and data continuity) and geographical location information , where is the latitude of the station, is the longitude of the station.

[0049] S13. Set the preset threshold to distinguish between stations with normal and abnormal data. For each station, if , it is determined as a low-health station, denoted as :

[0050] Here, is determined according to historical data statistics or expert experience.

[0051] Meanwhile, for the identified low-health stations, record their precise geographical coordinates for subsequent distance calculation.

[0052] Step S2: Screening of neighboring high-health stations and distance calculation. Screen out the stations with health reaching the predetermined standard within the set geographical radius (5 kilometers), and calculate the spatial distance between the target station and each neighboring station using the spherical distance formula; S21. Screening of neighboring stations. For each low-health station , set a fixed geographical radius around the target station to determine the candidate high-health stations. Define the candidate station set as:

[0053] where represents station and station The spatial distance between

[0054] S22. Spherical distance calculation (Haversine formula). For the distance between stations, the Haversine formula is used. This calculation takes into account the curvature of the earth, and the calculation method is as follows:

[0055] Where = 6371 km is the radius of the earth, is the difference in latitude between the two stations, is the difference in longitude between the two stations, and are the latitudes of stations respectively, and are the longitudes of stations respectively.

[0056] By converting the coordinates and the coordinates of the candidate high-health stations as follows:

[0057]

[0058]

[0059] Substituting the above interpolation into the Haversine formula, we get:

[0060] Obtaining the set corresponding to each station and its distance providing a data basis for subsequent weight calculation and IDW interpolation method; Where and are the converted radian coordinates of the latitude coordinates and respectively, and are the converted radian coordinates of the longitude coordinates and respectively.

[0061] At this time, by using the spherical distance calculation formula in the adjacent site screening stage to accurately describe the geospatial relationship between sites, and by utilizing the spatial correlation and data redundancy among multiple sites, it is possible to overcome the calculation error of traditional planar distance, improve the topographic adaptability of the interpolation result, ensure the accuracy and continuity of rainfall data repair, and further enhance the reliability of rainfall data repair under complex terrain conditions. It is applicable to fields such as flood forecasting, flood control dispatching, and hydrological monitoring.

[0062] Step S3: Weight calculation. According to the inverse distance weighting principle, calculate the contribution weight of each adjacent site to the data repair of the target site by using the spherical distance, power exponent, and a small constant. Refer to Figure 2 ; S31. For each high-score adjacent site , calculate the weight :

[0063] Among them, ( ) ensures that there will be no division-by-zero error when the site approaches zero. ( ) makes the closer sites have larger weights. is the distance between each site and its adjacent site . is the total number of high-score adjacent sites. is the starting number of the summation. is the weight adjustment value.

[0064] Step S4: Rainfall data repair. Adopt the IDW interpolation method to interpolate the missing or abnormal data of low-health sites with the rainfall data and weights of adjacent sites to generate continuous and high-precision rainfall time series data; S41. By using the IDW interpolation method, calculate the rainfall value of the target site at time according to the calculated weight ; The rainfall estimated value of the target site :

[0065] Among them, is the total number of high-score adjacent sites. is the rainfall estimated value of site at time .

[0066] At this time, by introducing the inverse distance weighted interpolation (IDW) method in the rainfall data repair process, the spatial correlation of neighboring high-health stations can be fully utilized, effectively making up for the limitations of traditional simple interpolation methods in data sparse areas, thereby significantly improving the data repair accuracy and continuity of low-health stations, and facilitating the provision of high-quality, high-temporal and spatial matching input data for flood forecasting models.

[0067] Through steps S3 and S4, the mathematical modeling of weight calculation and data repair is completed, ensuring that when the health of the rainfall station is low or the data is missing, the data of the neighboring stations can be used for high-precision interpolation to generate complete rainfall time series data. At this time, the comparison of the rainfall data time series before and after the repair is as follows: Figure 3 shown.

[0068] In summary, the integrity and reliability of rainfall data are significantly improved through time alignment, missing value filling and weighted fusion of neighboring stations, ensuring accurate matching of data between monitoring stations, thereby providing high-quality input data for flood forecasting models.

[0069] Step S5: Data verification and model integration: statistical tests and continuity assessments are performed on the repaired data to remove abnormal segments with unsatisfactory interpolation effects. The verified data are then fused with other hydrological monitoring data as input features for the flood forecasting model.

[0070] S51, set the target site in the time interval The repair data in , the corresponding statistical indicators include mean, variance and continuity error. For repair data, calculate the mean and variance :

[0071] in, is the length of the time interval; S52: To evaluate data continuity, the continuity error index is introduced :

[0072] When the continuity error When the set threshold value θ is exceeded, the interpolation effect of this time period is considered unsatisfactory and is eliminated.

[0073] At this time, by introducing statistical tests and continuity error assessment mechanisms in the data verification stage, abnormal data segments with unsatisfactory interpolation effects can be automatically eliminated, thereby ensuring the consistency and continuity of the repaired data in a statistical sense, and ultimately avoiding the risk of misjudgment of the flood forecasting model due to local interpolation deviations.

[0074] S53. Data fusion and model integration. After statistical tests and continuity evaluations, select the qualified repaired data Fuse it with other hydrological monitoring data (such as water level, flow rate, etc.) to construct a joint input feature vector . The fusion process is described as follows:

[0075] Where 、 represent other relevant monitoring data such as reservoir water level and flow rate respectively. After ensuring that each data is accurately matched in space and time, it is used as the input of the flood forecasting model represents the effective time period

[0076] At this time, by realizing the spatio-temporal fusion of multi-source hydrological data in the model integration stage, a joint input feature vector can be constructed to match the requirements of the flood forecasting model, so as to improve the dynamic response ability of the model to multi-dimensional hydrological processes, and finally enhance the spatio-temporal resolution and prediction accuracy of the flood forecasting results

[0077] Step S6: Model performance evaluation and optimization. Use indicators such as mean square error (MSE) to quantitatively evaluate the forecasting model, and optimize the data repair algorithm and model parameters based on the feedback to construct an adaptive early warning system

[0078] S61. In this step, use indicators such as mean square error (MSE) to quantitatively evaluate the output of the flood forecasting model

[0079] Let be the true observed rainfall data (or other hydrological indicators), be the model prediction value, then the mean square error is defined as:

[0080] S62. According to the calculated value, after comparing it with the preset performance indicators, tune the data repair algorithm (and the hyperparameters of the forecasting model. Use gradient descent to minimize the model output error:

[0081] Where represents the set of all parameters to be optimized

[0082] At this time, by introducing an adaptive adjustment mechanism driven by mean square error (MSE) in the model optimization stage, the data repair algorithm and model parameters can be dynamically optimized in real time, thereby enhancing the system's robustness to extreme climate events and emergencies, and facilitating the final construction of an intelligent flood control early warning system with adaptive capabilities

[0083] S63. Construction of an adaptive early warning system, which realizes the adaptive ability of the system through real-time monitoring and feedback adjustment of the model output error. Specifically, when exceeds a certain early warning threshold , the system automatically triggers the parameter re-optimization or alarm mechanism to timely respond to data anomalies and emergencies, ensuring the stability and accuracy of the overall forecasting system.

[0084] In summary, this method analyzes the spatial relationship between low-health stations and their neighboring high-health stations. Within the set geographical radius, by calculating the distances between low-health stations and each high-health station, different contribution weights are assigned to each neighboring station according to the inverse distance weighting principle, so as to interpolate and repair the missing or abnormal rainfall data of low-health stations, generate continuous and high-precision rainfall time series data, and after the repaired rainfall data is fused with other hydrological monitoring data (such as water level, flow, etc.), it is used as the input feature of the flood forecasting model. The performance of the forecasting model is evaluated by the mean square error (MSE) index, and the data repair algorithm and model parameters are optimized according to the evaluation results, thereby constructing an adaptive early warning system that can accurately predict the trend of flood occurrence. Moreover, this method makes full use of the spatial correlation and data redundancy between multiple stations, ensuring the accuracy and continuity of rainfall data repair, and providing scientific and reliable data support for flood control dispatching and water conservancy project management.

[0085] On the other hand, the present invention also discloses a rainfall data repair system for executing the method in the above embodiments, including: A station data acquisition module for collecting the original monitoring data of all rainfall stations in the target area and obtaining the health index and geographical location information; A data calculation and repair module for performing interpolation calculation on the missing or abnormal data of low-health stations based on the rainfall data and weights of neighboring stations to generate continuous and high-precision rainfall time series data; A model verification module for performing statistical tests and continuity evaluations on the repaired data, removing abnormal segments with unsatisfactory interpolation effects, and after fusing the verified data with other hydrological monitoring data, using it as the input feature of the flood forecasting model.

[0086] By the organic combination of spatial correlation modeling, multi-source data fusion and adaptive optimization technology, this system can break through the limitations of traditional data repair methods, thereby comprehensively improving the data quality, prediction accuracy and real-time response ability of the flood forecasting system, and ultimately providing scientific and reliable technical support for flood control dispatching, water conservancy project management and disaster prevention and control, and improving the overall accuracy and real-time response ability of the forecasting system.

[0087] In another aspect, the present invention also discloses a computer-readable storage medium storing a computer program, which when executed by a processor causes the processor to execute the steps of the above method.

[0088] In yet another aspect, the present invention also discloses a computer device including a memory and a processor, where the memory stores a computer program, which when executed by the processor causes the processor to execute the steps of the above method.

[0089] In yet another embodiment provided by the present application, there is also provided a computer program product containing instructions, which when running on a computer causes the computer to execute the overall method of any one of the above embodiments for rainfall station data repair and flood forecasting model integration based on inverse distance weighted interpolation.

[0090] It can be understood that the system provided by the embodiments of the present invention corresponds to the method provided by the embodiments of the present invention. Explanations, examples, and beneficial effects of related content can refer to the corresponding parts in the above method.

[0091] The embodiments of the present application also provide an electronic device including a processor, a communication interface, a memory, and a communication bus. Among them, the processor, the communication interface, and the memory complete communication with each other through the communication bus. The memory is used to store a computer program. The processor is used to implement the overall method of the above rainfall station data repair and flood forecasting model integration based on inverse distance weighted interpolation when executing the program stored on the memory.

[0092] The communication bus mentioned in the above electronic device may be a peripheral component interconnect standard bus or an extended industry standard architecture bus, etc. This communication bus can be divided into an address bus, a data bus, a control bus, etc.

[0093] The communication interface is used for communication between the above electronic device and other devices.

[0094] The memory may include a random access memory and may also include a non-volatile memory, such as at least one disk memory. Optionally, the memory may also be at least one storage device located far from the aforementioned processor.

[0095] The above-mentioned processor may be a general-purpose processor, including a central processing unit, a network processor, etc.; it may also be a digital signal processor, an application-specific integrated circuit, a field programmable gate array, or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0096] It should also be noted that the electronic device further includes a terminal device, which can also be referred to as a terminal, user equipment, mobile station, mobile terminal, etc. The terminal device can be a mobile phone, smart TV, wearable device, tablet computer, computer with wireless transceiver function, virtual reality terminal device, augmented reality terminal device, wireless terminal in industrial control, wireless terminal in unmanned driving, wireless terminal in remote surgery, wireless terminal in smart grid, wireless terminal in transportation safety, wireless terminal in smart city, wireless terminal in smart home, and so on. The specific technologies and device forms adopted by the terminal device in the embodiments of the present application are not limited.

[0097] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website, computer, server, or data center to another website, computer, server, or data center in a wired (such as coaxial cable, optical fiber, digital subscriber line) or wireless (such as infrared, wireless, microwave, etc.) manner. The computer-readable storage medium can be any available medium that can be accessed by a computer, or a data storage device such as a server or data center that includes one or more available media integrated. The available medium can be a magnetic medium (such as a floppy disk, hard disk, magnetic tape), an optical medium, or a semiconductor medium (such as a solid-state drive), etc.

[0098] The above are only the preferred embodiments of the present invention, and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection scope of the present invention.

[0099] In addition, it should be noted that if there are directional indications (such as up, down, left, right, front, back...) involved in the embodiments of the present invention, the directional indications are only used to explain the relative position relationship and movement conditions between components in a specific posture. If the specific posture changes, the directional indications will also change accordingly.

[0100] In addition, if the descriptions such as "first" and "second" are involved in the embodiments of the present invention, the descriptions of "first", "second", etc. are for descriptive purposes only, and cannot be construed as indicating or implying their relative importance or implicitly indicating the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one such feature. In addition, the meaning of "and / or" appearing throughout the text includes three parallel scenarios. Taking "A and / or B" as an example, it includes Scenario A, or Scenario B, or the scenario where both A and B are satisfied simultaneously. In addition, in the embodiments of the present invention, "a plurality of" means two or more. In addition, the technical solutions between the various embodiments can be combined with each other, but it must be based on the ability of those of ordinary skill in the art to implement. When the combination of technical solutions results in contradictions or cannot be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

Claims

1. A method for repairing rainfall data, characterized in that: The following steps are involved: S1. Read all rainfall stations in the target area ( , is the number of all rainfall stations in the target area), recorded as a time series ,in Indicates at time The rainfall value at the moment, and the health index of each station and geographic location information coordinates ,in is the latitude of the site, is the longitude of the site, and based on the preset threshold Screen out low-health sites with abnormal data or monitoring failure , determine its precise spatial coordinates; S2. Select the sites whose health reaches the predetermined standard within the preset geographic radius, and calculate the spatial distance between the target site and each neighboring site using the spherical distance formula; S3. According to the inverse distance weighting principle, the contribution weight of each neighboring site to the target site data restoration is calculated using the spherical distance, power exponent and small constant; S4. The IDW interpolation method is used to interpolate missing or abnormal data of low-health stations using the rainfall data of neighboring stations and their weights to generate continuous and high-precision rainfall time series data; S5. Perform statistical tests and continuity assessments on the data and remove abnormal segments with unsatisfactory interpolation effects.

2. The rainfall data repair method according to claim 1, characterized in that: The specific operation process of the S2 step includes: S21. Neighboring site screening, for each low health site , set a fixed geographic radius around the target site , used to determine candidate high health sites, and define the candidate site set as: in, Indicates the site With site The spatial distance between S22. Calculate the distance between sites. The calculation method is as follows: in is the radius of the Earth, is the difference in latitude between the two stations, is the difference in longitude between the two stations, and Site The latitude and longitude of and Site latitude and longitude; The coordinates With candidate high health sites Coordinates For conversion, the interpolation calculation formula is as follows: Substituting the above interpolation into the distance calculation formula between stations, we have: To get the collection of each site The corresponding set and its distance ; in, and The latitude coordinates are and The converted radian coordinates, and Longitude coordinates and The converted radian coordinates.

3. The rainfall data repair method according to claim 1, characterized in that: The steps S3 and S4 are used to calculate weights and generate continuous and high-precision rainfall time series data, including: For each high-scoring neighboring site , calculate the weight , the calculation formula is: in, Used to ensure that division by zero does not occur when the station approaches zero. This makes the closer sites have greater weight. For each site Its neighboring sites The distance between is the total number of high-scoring neighboring sites, To sum the starting number, is the weight adjustment value; By using the IDW interpolation method, according to the calculated weights Calculate target site In time Estimated rainfall at ,have: It is used to generate complete rainfall time series data by using the data of neighboring stations for high-precision interpolation when the health of the rainfall station is low or the data is missing. is the total number of high-scoring neighboring sites, For Site In time Estimated rainfall at .

4. The rainfall data repair method according to claim 1, characterized in that: The specific operation process of step S5 includes: S51. Set target site In the time interval Repair data within for , at this time, for the repair data, calculate the mean and variance , the calculation formula is: in, is the length of the time interval; S52. Set the continuity error index , used to evaluate the data continuity error index The calculation formula is: When the continuity error Exceeding the preset threshold , remove this time interval Interpolation within .

5. A rainfall data repair system, used to execute the rainfall data repair method according to any one of claims 1 to 4, characterized in that: include: The site data collection module is used to collect the original monitoring data of all rainfall sites in the target area and obtain health indicators and geographical location information; The data calculation and repair module is used to interpolate missing or abnormal data of low-health stations based on the rainfall data of neighboring stations and their weights to generate continuous and high-precision rainfall time series data; The model verification module is used to perform statistical tests and continuity assessments on the repaired data, remove abnormal segments with unsatisfactory interpolation effects, and fuse the verified data with other hydrological monitoring data as input features for the flood forecasting model.

6. A flood forecasting method, characterized in that: Based on the rainfall data repair method described in any one of claims 1 to 4 above, the repaired rainfall data is obtained, and the following specific operation steps are included: L1. Based on the estimated rainfall at each station , construct a flood forecasting model, and fuse the verified data with other hydrological monitoring data as input features of the flood forecasting model; L2. Use preset indicators including mean square error to quantitatively evaluate the forecast model, optimize the data repair algorithm and model parameters based on feedback, and build an adaptive early warning system; L3. Carry out flood prediction operations through flood forecasting models.

7. The flood forecasting method according to claim 6, characterized in that: The fusion method of input features in the L1 step includes: Select repair data that pass statistical tests and continuity assessments , fused with other hydrological monitoring data to construct a joint input feature vector ,have: in, , Represent reservoir water level data and flow data respectively, Indicates the valid time period.

8. The flood forecasting method according to claim 6, characterized in that: The specific optimization process in the L2 step includes: L21. Use specified indicators including mean square error to quantitatively evaluate the output of flood forecasting models. Within, there is a mean square error The quantitative evaluation is defined as: in, Hydrological indicators including real observed rainfall data, is the model prediction value; L22. According to the calculation After comparing the value with the preset performance index, the hyperparameters of the data repair algorithm and the prediction model are optimized, and the gradient descent is used to minimize the model output error. ,in Represents the set of all parameters to be optimized; L23. Build an adaptive early warning system to monitor and adjust the model output error in real time. Exceeds the specified warning threshold When the system automatically triggers parameter re-optimization or alarm mechanism.

9. A computer-readable storage medium, characterized in that: A computer program is stored, and when the computer program is executed by a processor, the processor executes the steps of the method according to any one of claims 1 to 4, or executes the steps of the method according to any one of claims 6 to 8.

10. A computer device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the method as claimed in any one of claims 1 to 4, or executes the steps of the method as claimed in any one of claims 6 to 8.

Images